Skeletal Radiol (1995) 24:111-115

Skeletal Radiology

Bone island (enostosis): current c o n c e p t - a review

Adam Greenspan, M.D.

Department of Radiology, University of California, Davis School of Medicine, Sacramento, California, USA

Abstract. An enostosis or bone island represents a focus of mature compact (cortical) bone within the cancellous bone (spongiosa). Thought by some to be a tumor-like condition and by others a hamartoma, this benign lesion is probably congenital or developmental in origin and re- flects failure of resorption during endochondral ossifica- tion. A bone island can be virtually diagnosed based on its characteristic clinical and radiologic features. Typi- cally asymptomatic, the lesion is usually an incidental finding, with a preference for the pelvis, femur, and oth- er long bones, although it may be found anywhere in the skeleton, including the spine. Plain radiography reveals a homogeneously dense, sclerotic focus in the cancellous bone with distinctive radiating bony streaks ("thorny ra- diation") that blend with the trabeculae of the host bone, creating a feathered or brush-like border. On CT scan, a bone island appears as a low-attenuation focus, and on MRI sequences it shows low signal intensity like cortical bone. A distinguishing feature of bone islands is that they are usually "cold" on skeletal scintigraphy. Thus, bone scan has been and continues to be the means of dif- ferentiating bone islands from the more aggressive enti- ties. However, reports of histologically confirmed bone islands that were scintigraphically active have raised a note of caution about relying on this modality in the dif- ferential consideration of lesions otherwise characteristic of bone islands. Guides to the correct diagnosis should be looked for in the individual clinical situation and in the morphologic features of the lesion on plain radiogra- phy, CT, and MRI, without regard to the lesion's activity on bone scan. If such a lesion, however, is symptomatic and "hot" on scintigraphy, it demands close observation with follow-up imaging studies.

Key words: Developmental anomalies - Bone, sclerosis - Bone island - Enostosis - Scintigraphy

Correspondence to: Adam Greenspan, M.D., Department of Dia- gnostic Radiology, University of California, Davis Medical Center, 2516 Stockton Blvd., Ticon II, Sacramento, CA 95817, USA

Since its earliest descriptions in the literature, first by Stieda [1] in 1905 and 7 years later by Fischer [2], an enostosis or bone island has been variously named and defined. Stieda referred to the small, dense, circum- scribed shadows he observed inside the cancellous por- tion of short tubular bones and in the articular ends of long bones as "compact bone nuclei" (kompakte Kno- chenkerne). It was Fischer, however, who described these lesions as compact "islands" and emphasized their importance in differential diagnosis [2]. Others have of- fered further names, including "calcified island in med- ullary bone" by Steel [3], "sclerotic bone island" by Meschan [4], "focal sclerosis" by Caffey [5], and "end- osteoma" by Schmorl and Junghanns [6] (describing ver- tebral lesions). Definitions of this entitiy, as the names imply, have been similarly varied. While some still clas- sify enostosis among tumor-like conditions [7-9], Mirra regards it as "misplaced, hamartomatous cortical bone" [10]. In addition, recent investigation suggests that foci of mature compact bone within the spongiosa, character- istic of this condition, represent areas that failed to re- sorb during endochondral ossification [11]; probably de- velopmental or congenital in origin, enostosis is thus an anomaly apparently closely related to osteopoikilosis [11-13]. The importance of recognizing this benign le- sion, which can be virtually diagnosed on the basis of its characteristic clinical and radiologic features, consists in the need to differentiate it from clinically more signifi- cant bone lesions such as primary or metastatic tumors when it manifests itself uncharacteristically by being very large [14] and showing activity on skeletal scintig- raphy [12].

Clinical presentation

Typically asymptomatic, a bone island is often an inci- dental finding on radiography performed for another purpose. It is discovered more frequently in adults than in children and shows no sexual predilection. The pelvis, the femur, and other long bones are preferred sites of in-

�9 1995 International Skeletal Society

112 A. Greenspan: Bone island

volvement, although the lesion may be found anywhere in the skeleton, including the carpal and tarsal bones and the ribs [15]. The spine is a rare site of involvement [6, 16-18], accounting for only three (1.4%) of 209 bone is- lands reviewed by Onitsuka [19]; these vertebral bone is- lands involved the thoracic and lumbar segments [19].

Imaging

Plain-film radiography

A bone island exhibits a consistent radiographic picture regardless of its site or size. The lesion appears as an ovoid, round, or oblong focus, homogeneously dense and sclerotic, in the cancellous bone (spongiosa). It is usually oriented with the long axis of the bone parallel to the cortex. Highly distinctive of this lesion are radiat- ing bony streaks, aligned with the axis of the host bone's trabeculae, that merge with the trabeculae in a feathered or brush-like fashion (Fig. 1). This "thorny radiation" [20, 21] or "pseudopodia" [12] can be better demonstrat- ed with conventional tomography (Fig. 2). Bone islands in the great majority of cases range in size from 1 mm to 2 cm [5, 14]. "Giant" bone islands, defined as lesions larger than 2 cm [22], have also been reported [14]. The largest lesion on record, described by Gold and co-work- ers [22], measured 5x5x4.5 cm and was located in the tibia. Other giant enostoses have been observed by Smith [14] (3.5x4 cm, located in the ilium) and Sickles and colleagues [23] (3 cm, located in the femur). In ad-

dition to showing the same radiographic features as smaller lesions, giant bone islands may also give the im- pression of "cumulus cloud" [10].

Bone islands usually do not show radiographic evi- dence of change in size over time. Nonetheless, several investigators have reported size changes in bone islands, some of which exhibited metabolic activity [15, 19, 24-27]. In the large series of patients who had multiple examinations described by Onitsuka [19], 44 of 138 bone islands (31.9%) showed changes in radiographic size over periods ranging from 3 to 23 years. After ex- clusion of 19 of these lesions because they had propor- tionally enlarged during adolescent growth, 21 of the re- mainder had increased in size, and 4 had became small- er. Other evidence of growing bone islands is offered by Blank and Lieber [24], who reported six cases, by Ngan [15], who described three cases and emphasized the pos- sible confusion of such lesions with sclerotic metastases, and by Hoffman and Campbell [26], who reported the unstable appearance of a bone island. In the latter work the instability of a bone island was demonstrated by its disappearance and reappearance over a period of 6 years in a patient with hyperparathyroidism.

CT and MRI

A bone island appears as a low-attenuation focus on CT, exhibiting (as in plain films) its characteristic brush bor- ders (Fig. 3A). Occasionally the lesion's "pseudopodia" may show more rounded contours (Fig. 3B). On both

Fig. 1 A, B. Plain-film radiography of bone island. A Sclerotic focus in the spongiosa with distinctive radiating streaks that are aligned with the axis of the trabeculae of the host bone. B Radiograph of the femoral-head specimen shows the bone island's "thorny radiation" that blends with the trabeculae of the host bone. (B Courtesy of German Steiner, M.D., New York, NY, USA)

Fig. 2. Trispiral tomography demonstrates a bone island in the posterior aspect of the calcaneus. Note the "brush" borders and elongated "pseudopodia"

Fig. 3 A, B. CT of bone isl.and. A Low-attenuation focus in the medial femoral condyle displays sharp radiating bony streaks. B Low-attenuation focus in the medial aspect of tibia shows "pseudopodia" with more rounded contours

A. Greenspan: Bone island 113

linked to the large size of reported bone islands [23] and to their growth, based on the observation that uptake of radiopharmaceutical is roughly proportional to the vol- ume of a lesion [23]. Yet two large bone islands that were undetectable on bone scans, as reported by Hall and co-workers [29], indicate that a lesion's size alone is not always a factor in this phenomenon. Others have pointed to a giant bone island's histologic morphology and increased metabolic activity relative to the surround- ing cancellous bone as the probable explanation for its striking radionuclide activity [22]. Because tracer uptake implies greater metabolic activity and, usually, increased blood flow, it is logical to assume that a bone island will show increased activity if it becomes metabolically ac- tive. However, reports of radiographically documented growing bone islands (one with a positive and the other with a negative bone scan) that showed no histologic ev- idence of metabolic activity raise doubts about this pathomechanism [24, 29]. A recent contribution by Greenspan and co-workers [12], correlating the radiolog- ic and pathologic findings in six cases of bone island, suggested that the increased tracer uptake observed in two bone islands appeared to be directly related to the higher degree of bone remodeling and osteoblastic activ- ity that these lesions exhibited in comparison with the scintigraphically "cold" lesions.

Fig. 4 A, B. MRI of bone island. A Coronal spin-echo T1- weighted MR image shows a low-signal-intensity giant lesion in the medial femoral condyle. B On coronal spin-echo T2-weigbted MR image the lesion continues to show low signal intensity

Fig. 5 A, B. Scintigraphic image of "hot" bone island. A Plain- film radiograph shows a bone island in the medial aspect of the left tibia in a patient with breast carcinoma. B After administration of 15 mCi (555 MBq) of 99mTc-labeled methylene diphosphonate, the lesion shows an increased uptake of radiotracer. (Reprinted with permission from [12])

T1- and T2- (or T2*-)weighted MRI sequences, a bone island exhibits the low signal intensity characteristic of cortical bone (Fig. 4).

Scintigraphy

A distinctive feature of bone islands is that for the most part they show no activity on skeletal scintigraphy, a finding generally ascribed to their having a level of met- abolic activity about the same as that of the surrounding cancellous bone. Thus, radionuclide imaging has been and continues to be the means of differentiating bone is- lands from more aggressive bone lesions. Nonetheless, reports in the literature of at least nine histologically confirmed bone islands that showed increased radiotrac- er uptake on bone scan [12, 22, 23, 28, 29] have raised a note of caution about the usefulness of scintigraphy in distinguishing a bone island from clinically more signifi- cant lesions (Fig. 5).

The phenomenon of scintigraphically active bone is- lands has prompted a number of explanations of the pathomechanism, which is still unclear. It has been

Histopathology

Microscopic examination of a bone island reveals a his- tologic picture that correlates with the radiologic find- ings. The lesion is a focus of compact (cortical) bone in the spongiosa (Fig. 6A,B) that shows a mature lamellar configuration and a surrounding haversian system of nu- trient canals (Fig. 6C). It is occasionally connected to the endocortex of the host bone [10, 30]. In addition, thorn-like, thickened trabeculae radiating from the lesion merge with the surrounding trabeculae of the spongiosa. Bone islands occasionally contain foci of woven, nonla- mellar bone, and, as reported in some scintigraphically active bone islands, they may consist of a mixture of la- mellar and woven bone [12, 22]. No or only a very few osteoblasts and osteoclasts are usually observed in bone islands; foci of osteoblastic and osteoclastic activity are rare.

Differential diagnosis

In addition to its inclusion in the differential diagnosis of calcifying enchondroma and medullary bone infarct as well as a healing nonossifying fibroma, which may occa- sionally mimic it, an enostosis should be considered in the differential diagnosis of sclerotic medullary lesions that carry far more significant clinical implications than bone island, including osteoid osteoma, osteoblastoma, osteosarcoma, and sclerotic metastases. Scintigraphic activity in a lesion such as an enostosis that is generally thought to be "cold" on bone scan raises the possibility that a benign entity will be misdiagnosed as a more seri-

114 A. Greenspan: Bone island

Fig. 6 A-C. Histopathology of bone island. A Low-power photomicrograph of the femoral head (H&E, original magnifi- cation x4) shows a focus of cortical bone within spongiosa (arrow). (Courtesy of German Steiner, M.D., New York, NY, USA). B Low-power photomicrograph of bone island (H&E, original magnification x16) shows multiple haversian nutrient canals in

the center of the lesion. Note the characteristic "pseudopodia." (Reprinted with permission from [11]. C High-power photo- micrograph (H&E, original magnification x90) shows compact bone with wide bands of parallel lamellae surrounding nutrient canals. (Courtesy of German Steiner, M.D., New York, NY, USA)

ous abnormality. Of course, the possibility also exists, regardless of activity on radionuclide imaging, that a tu- mor might be misdiagnosed as a bone island [10]. De- spite the continuing usefulness of bone scintigraphy in distinguishing bone islands from more aggressive le- sions, finding activity in an otherwise suspected bone is- land cannot be a valid factor in the differential consider- ations [12].

In general, the presumptive diagnosis of a bone island can easily be made if the individual clinical and radio- logic findings, together with follow-up examinations, are taken into account. The guides to the correct diagnosis are found in the lesions's morphologic features as dem- onstrated on plain radiography, CT, and MRI, without reliance on scintigraphy findings. Thus, in a patient found to have an asymptomatic isolated sclerotic bone lesion showing the typical features of a bone island, with brush or feathered borders, the most likely diagnosis is an enostosis, whatever its size or its activity on bone scan. Even the discovery of a sclerotic lesion in a patient with a known neoplasm would strongly suggest a bone island if this lesion's characteristic features are observed and bone scan is normal. Whenever a patient is symp- tomatic, however, or the lesion appears "hot" on scintig- raphy, misdiagnosis may occur in difficult cases. Such circumstances would demand careful observation with follow-up imaging studies. An open biopsy is in order if, as Mirra [10] suggested, the lesion's growth exceeds

25% of its diameter within 6 months or 50% within 1 year.

Conclus ion

A bone island or enostosis is a focus of mature compact (cortical) bone within the cancellous bone (spongiosa), reflecting a developmental error during the process of endochondral ossification. The lesion is typically asymptomatic, manifests itself with highly distinctive ra- diologic features, and can be virtually diagnosed on these grounds. The usual inactivity of enostosis on scin- tigraphy has been considered a discriminating feature, and thus a radionuclide bone scan continues to be the means of distinguishing bone islands from the more ag- gressive entities that should be included in the differen- tial diagnosis of sclerotic medullary lesions, including osteoid osteoma, osteoblastoma, osteosarcoma, and os- teoblastic metastasis. However, reports of histologically confirmed bone islands that showed activity on bone scans have raised a note of caution about relying on scintigraphy when a lesion otherwise characteristic of an enostosis is "hot". The individual clinical circumstances and the morphologic features of the lesion on plain radi- ography, CT, and MRI should guide the differential con- siderations.

A. Greenspan: Bone island 115

Acknowledgements. I would like to thank Deborah Ann Hoang for effective secretarial assistance in the preparation of this manu- script.

References

1. Stieda A. Ueber umschriebene Knochenverdichtungen in Be- reich der Substantia spongiosa in Roentgenbildern. Bruns Beitr Klin Chir Tuebing 1905; 45: 700.

2. Fischer H. Contributions to information concerning variations of the skeleton. Fortschr Geb Roentgenstr 1912; 19: 43.

3. Steel HH. Calcified islands in medullary bone. J Bone Joint Surg [Am] 1950; 32-A: 405.

4. Meschan I. Normal variant sclerotic bone island. In: Mechan I, ed. Roentgen signs in clinical diagnosis. Philadelphia; Saun- ders, 1957: 256.

5. Caffey J. Focal sclerosis of spongiosa. In: Caffey J, ed. Pediat- ric X-ray diagnosis. Chicago: Year Book Medical, 1961: 842.

6. Schmorl G, Junghanns H. The human spine in health and dis- ease, 2nd edn. New York: Grune and Stratton 1971: 327.

7. Hudson TM. Radiologic-pathologic correlation of musculo- skeletal lesions. Baltimore: Williams & Wilkins, 1987: 31-33.

8. Resnick D, Kyriakos M, Greenway GD. Tumors and tumor- like lesions of bone: imaging and pathology of specific le- sions. In: Resnick D, ed. Bone and joint imaging. Philadel- phia: Saunders 1989:1107.

9. Schajowicz F, Ackerman LV, Sissons HA. Histological typing of bone tumors. International histological classification of tu- mors. Geneva: World Health Organization, 1972.

10. Mirra JM, Picci R Gold RH. Bone tumors: clinical, radiologic and pathologic correlations. Philadelphia: Lea & Febiger 1989: 182.

11. Greenspan A. Sclerosing bone dysplasias - a target-site ap- proach. Skeletal Radiol 1991; 20: 561.

12. Greenspan A, Steiner G, Knutzon R. Bone island (enostosis): clinical significance and radiologic and pathologic correla- tions. Skeletal Radiol 1991; 20: 85.

13. Greenspan A. Sclerosing bone dysplasias. In: Taveras JM, Ferrucci JT, eds. Radiology - diagnosis, imaging, interven- tion, vol 5. Philadelphia; Lippincott 1993: 1-18.

14. Smith J. Giant bone islands. Radiology 1973; 107: 35. 15. Ngan H. Growing bone islands. Clin Radiol 1972; 23: 199. 16. Broderick TW, Resnick D, Georgen TG, Alazraki N. Enosto-

sis of the spine. Spine 1978; 3: 167. 17. Epstein BS. The spine: a radiological text and atlas, 3rd edn.

Philadelphia: Lea & Febiger, 1969. 18. Resnick D, Nemcek AA Jr, Haghighi R Spinal enostoses

(bone islands). Radiology 1983; 147: 373. 19. Onitsuka H. Roentgenologic aspects of bone islands. Radiolo-

gy 1977; 123: 607. 20. Kim SK, Barry WE Bone islands. Radiology 1968; 99: 77. 21. Kim SK, Barry WE Bone island. AJR 1964; 92: 1301. 22. Gold RH, Mirra JM, Remotti F, Pignatti G. Case report 527.

Skeletal Radiol 1989; 18: 129. 23. Sickles EA, Genant HK, Hoffer PB. Increased localization of

99mTc-pyrophosphate in a bone island: case report. J Nucl Med 1976; 17: 113.

24. Blank N, Lieber A. The significance of growing bone islands. Radiology 1985; 85: 508.

25. Go RT, E1-Khoury GY, Wehbe MA. Radionuclide bone image in growing and stable bone island. Skeletal Radiol 1980; 5: 15.

26. Hoffman RR Jr, Campbell RE. Roentgenologic bone island in- stability in hyperparathyroidism. Radiology 1972; 103: 307.

27. Scheel W, Strauss B, Eger H. Kostale Kompaktainseln in der Differentialdiagnostik kleiner Rundherde of Thoraxtibe- rsichtsaufnahmen. Z Erkrank Atm Org 1985; 164: 299.

28. Davies JAK, Hall FM, Goldberg RR Kasdon EJ. Positive bone scan in a bone island. J Bone Joint Surg [Aml 1979; 61-A: 943.

29. Hall FM, Goldberg RP, Davies JAK, Fainsinger MH. Scinti- graphic assessment of bone islands. Radiology 1980; 135: 737.

30. Lagier R, Nussle D. Anatomy and radiology of a bone island. Fortschr Roentgenstr 1978; 128: 261.

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具

http://www.xuebalib.com/cloud/http://www.xuebalib.com/http://www.xuebalib.com/cloud/http://www.xuebalib.com/http://www.xuebalib.com/vip.htmlhttp://www.xuebalib.com/db.phphttp://www.xuebalib.com/zixun/2014-08-15/44.htmlhttp://www.xuebalib.com/